A dynamic model between central aortic pressure and radial photoplethysmogram: Experimental proof of concept

A parametric model is proposed to predict the amplitude-normalized central aortic pressure (CAP) waveform from the radial photoplethysmogram (PPG). The structure of the model is linear autoregressive with exogenous input (ARX [2 2 0]). Results show that a second order model can predict the radial PPG from the normalized-CAP with a fitness value above 88%. One application of this technique is to accurately estimate the amplitude-normalized CAP waveform from radial PPG, with the possibility of determination of important vascular indices. The advantage of the radial artery as recording site compared to the finger is it being unaffected by the autoregulation mechanism.